Comments by "JBird" (@jbird4478) on "Some bad code just broke a billion Windows machines" video.

What bugs me is that such critical organisations apparently can't simply pull a system snapshot restore via network boot.
95
@NickAc I mean... we can reasonably assume there is hash checking, right? So that means the file was corrupted before its hash was calculated.
37
And it's not even on some specific edge case. It's all computers running Windows and receiving this update. That means they have not tested it even once.
4
@someguy4915 There's basically three options. 1) every laptop runs on the same or one of a few system images, 2) every laptop stores its own restore point and only pulls the command/code to restore it, or 3) there is one stored per laptop but they are kept in sync using only diffs. And why does that have to happen within a minute or two? Every device can simply pick a random time of the day to create a snapshot. On a sidenote: the internet is somewhat bigger than you seem to think. The world produces around 500 exabytes, or 500.000 TB per day.
1
@tma2001 If the file was corrupted early on the clients would have all checked against an invalid hash, which would pass successfully. A hash just checks if the file received is the same as the one that was sent.
1
@jurajkovac8507 Apparently you don't.
1
And the former CTO of Crowdstrike is now on the Homeland Security Advisory Board, and the Cyber Safety Review Board. As a famous author once wrote "the difference between fiction and reality is that fiction has to make sense."
1